Next up I have to rotate the center of the gantry so that the bars are one above the other. Two reasons:
1. They will keep the pulleys at either end from bending inwards, which loosens the belt and causes movement problems.
2. When the end effector in the center is vertical, it will be much easier to make a system for lifting the nozzle. My first two lifting systems were not good enough, so this change doesn’t create new work for me.
Edit:
The nozzle works, the pump works, and it all moves via IRC. The new nozzles from @Aidan Leitch solve an earlier problem with hard plastic nozzles that had no grip or compliance. Yay! Great video, Aidan!
Of course the source of most problems is other solutions. With this new nozzle I’m forced to design a new method to raise and lower the nozzle onto the pieces. This first draft
was a strong start. Sadly it let air leak in from several places and over time it started to slouch on one side. Later it occured to me that if the nozzle tries to pick up a piece attached to other pieces, it would probably pick them ALL up. So there needs to be some kind of ring around the nozzle that keeps other pieces on the table. Maybe also a software rule that says sideways movements are forbidden when the nozzle is down.
@Dan Royer enlisted my help in making a soft suction nozzle for the JigSolve. I utilized many of the same techniques and experiences from my soft robotics projects. The result was four nozzles with varying cup sizes. These were the outer diameters of the cups:
Each one was 20mm tall. They were casted in 3-part 3D printed molds with Ecoflex 00-50 silicone, but almost any RTV silicone would have worked. All nozzles came out quite nicely and the gripping strength scales roughly with the cup size. I made an Instructables guide and video detailing how I made the nozzles:
Instructables Guide: http://www.instructables.com/id/Silicone-Suction-Grippers/
Video:
Modeled this brass connector today. The threaded NTP 1/8 end will go on @Aidan Leitch‘s nozzle, and the barbed end will go on the silicone hose seen in the previous post.
Around this nozzle I’m 3D printing a 2 part block that will hold the brass piece on a servo. the rotation axis of the servo goes right through the center of the nozzle, which should make rotating jigsaw pieces easy.
Next I need a new way to raise & lower the servo/nozzle to pick up pieces.
If the servo proves to be ineffective then I’ll copy the work of @anthony.webb and use a stepper motor with a hollow shaft. Changing to a stepper motor will have a trickle down effect that changes so…so many things. Let’s hope that doesn’t happen!
I’ve been working on a robot called JigSolve that will finish my jigsaw puzzles. Think “Twitch Plays Pokemon” taken to the next level. I made great progress until I got to the nozzle, which is proving delightfully challenging.
I wrote a whole post about recent adventures and the lessons learned. Read it on Hackaday.io project page.
